Institutional Repository of Institute of Process Engineering, CAS (IPE-IR）
Not a member yet
40778 research outputs found
Sort by
Review of explainable machine learning for anaerobic digestion
No full textAnaerobic digestion (AD) is a promising technology for recovering value-added resources from organic waste, thus achieving sustainable waste management. The performance of AD is dictated by a variety of factors including system design and operating conditions. This necessitates developing suitable modelling and optimi-zation tools to quantify its off-design performance, where the application of machine learning (ML) and soft computing approaches have received increasing attention. Here, we succinctly reviewed the latest progress in black-box ML approaches for AD modelling with a thrust on global and local model interpretability metrics (e.g., Shapley values, partial dependence analysis, permutation feature importance). Categorical applications of the ML and soft computing approaches such as what-if scenario analysis, fault detection in AD systems, long-term operation prediction, and integration of ML with life cycle assessment are discussed. Finally, the research gaps and scopes for future work are summarized
Facile synthesis of N-rich carbon nanosheets derived from antibiotic mycelial dregs as efficient catalysts for peroxymonosulfate activation
No full textNitrogen-rich carbon nanosheets using hazardous waste penicillin mycelial dregs (PMD) as the sole precursor were successfully synthesized amid molten salt-assisted pyrolysis and applied as catalysts for peroxymonosulfate (PMS) activation to degrade acid orange 7 (AO7). The structural properties and the associated catalytic per-formances of carbon nanosheets were precisely regulated by molten salt (NaCl/KCl) mass ratios and pyrolysis temperatures. Carbon nanosheets prepared at a molten salt ratio of 6 and a pyrolysis temperature of 800 degrees C possessed an optimized catalytic performance, achieving both high effective and efficient decolorization of AO7 than activated carbon catalysts. It was likely attributed to the combination of high graphitic N content and defective carbon structures from nanosheets through quantitative structure-activity relationships analysis. Both radical and non-radical pathways were recognized to be responsible for AO7 degradation, while surface-bound radicals generated from catalyst surface-PMS complexes in non-radical pathways were the main reactive oxygen species. This work offers a green and facile method to prepare high graphitic N content and defect-rich carbon nanosheets from nitrogen-rich biowastes, highlighting its promising catalytic properties for environmental remediation, synchronously expanding the means of resourceful and harmless treatment of PMD to improve the sustainability of antibiotic pharmaceutical production
Migration and emission characteristics of trace elements in coal-fired power plant under deep peak load regulation
No full textThe trace elements (TEs) have caused great harm to the environment due to the large consumption of coal, and their emission from the coal-fired power plant (CFPP) has become a hot issue. The deep peak load regulation (DPLR) become a trend in the CFPP, which will affect the migration and emission of TEs. To explore the effect of the DLPR on the migration and emission characteristics of typical TEs in a 330 MW CFPP, the TEs field tests were carried out during the regulation period. Results showed that a higher load enhanced the migration of Pb, Mn, and Cr from bottom ash to fly ash, while it had little effect on the other TEs. More importantly, >99 % of TEs (93 % of Se) could be captured by air pollution control devices (APCDs), and the emission risk of Se and Mn increased with the load. Compared with the other TEs, it is particularly noteworthy that Se has a higher gaseous proportion in the flue gas, and the emission factor sharply increased from 165 MW to 297 MW. In addition, part of the particulate selenium transformed into a gaseous state across the ESP. This work contributes to understanding the migration characteristic of TEs during the DPLR process of CFPP and provides guidance for TEs control in the CFPP
A deep learning approach using temporal-spatial data of computational fluid dynamics for fast property prediction of gas-solid fluidized bed
No full textTo deal with the critical issue of long computational time in practical application of computational fluid dynamics (CFD), this paper presents a new approach of deep learning for voidage prediction (DeepVP) that couples short time CFD simulations (limited CFD iterations) with the deep learning method to accelerate the 2D voidage distribution prediction for a gas-solid fluidized bed at steady state. Short time CFD simulations are first performed to obtain a sequence of voidage distribution images containing the temporal-spatial property of a gas-solid fluidized bed of the early period. A deep learning model is built to predict the voidage distribution at steady state, which is achieved by implementing multi-scale convolutional neural networks based on the sequence of voidage images. The case study results for a bubbling bed show that the voidage distribution at steady state for the bubbling bed can be predicted with comparable accuracy of conventional CFD simulations at about 1/30th computational cost. Moreover, the DeepVP method exhibits better extrapolation capability than the deep learning approach merely based on CFD condition parameters
Role of mesoscale structure in gas-solid fluidization: Comparison between continuum and discrete approaches
No full textThe coarse-grained discrete particle model (DPM) is fast growing into a powerful tool and a useful counterpart of the widely used two-fluid model (TFM) in simulation of large-scale reactors. This work aims to study the role of mesoscale modeling in both TFM and DPM approaches to understand the advantage and disadvantage of each approach for further development. Both simulation approaches with and without considering mesoscale structures in drag modeling are systematically investigated through simulations of an industrial diameter-transformed fluidized bed reactor with complex reactions. It is found that considering mesoscale drag can obviously improve the prediction in solid concentration for both approaches, and the effect of mesoscale drag for TFM modeling is more significant than for DPM approach. Besides, the DPM approach can reveal local heterogeneous structures without using mesoscale drag because it can distinguish different parcels in each fluid cell, but it overestimates the accumulation of solid particles below the distributor, as the large coarse-grain ratio may over-enhance the particle collision. For reaction, the coke content can be better predicted by both approaches with mesoscale drag, and the DPM simulation can capture more heterogeneous distribution of coke content than TFM modeling. The predicted temperature and product distribution still have obvious deviation from industrial data, suggesting a need of mesoscale heat and mass transfer modeling. The underlying mechanisms are further analyzed together with proposing future work
Sn-Ag Synergistic Effect Enhances High-Rate Electrocatalytic CO2-to-Formate Conversion on Porous Poly(Ionic Liquid) Support
No full textThe electrocatalytic transformation of carbon dioxide (CO2) to formate is a promising route for highly efficient conversion and utilization of CO2 gas, due to the low production cost and the ease of storage of formate. In this work, porous poly(ionic liquid) (PPIL)-based tin-silver (Sn-Ag) bimetallic hybrids (PPILm-SnxAg10-x) are prepared for high-performance formate electrolytic generation. Under optimal conditions, an excellent formate Faradaic efficiency of 95.5% with a high partial current density of 214.9 mA cm(-2) is obtained at -1.03 V (vs reversible hydrogen electrode). Meanwhile, the high selectivity of formate (>approximate to 83%) is maintained in a wide potential range (>630 mV). Mechanistic studies demonstrate that the presence of Ag-species is vital for the formation, maintenance, and high dispersion of tetravalent Sn(IV)-species, which accounts for the active sites for CO2-to-formate conversion. Further, the introduction of Ag-species significantly enhances the activity by increasing the electron density near the Fermi energy level
